Study of gene polymorphisms is increasingly used for diagnosing human diseases, assessing the risk of human diseases (e.g., finding atherosclerosis susceptibility genes), monitoring therapeutic interventions (e.g., detecting minimal residual disease in cancer or chimerism after transplantation), and predicting drug response (pharmacogenomics). Recently, whole genome amplification techniques through which minute amounts of DNA can be multiplied to generate quantities suitable for genetic testing and analysis (e.g., genotyping and DNA sequencing) have become available. We began studying the analytical performance and diagnostic reliability of one such commercial technique (REPLI-g, Qiagen) that is claimed to be applicable to purified genomic DNA, fresh or dried blood, buffy coat, and other sources of cells (e.g., buccal or tissue culture cells). This isothermal (i.e., non-thermal cycling) technique is based on so-called multiple displacement amplification. The key component of multiple displacement amplification is a uniquely processive DNA polymerase (phage 29) capable of activating very large (up to 100 kilobase) DNA segments without dissociating from the genomic template. Our initial results indicate that this whole genome amplification technique provides a uniform yield of high-molecular-weight DNA that is suitable for genetic studies. In another work, we reviewed currently available state-of-the art methods and future directions for nucleic acid-based diagnostic testing. In a collaborative study, we studied cerebrospinal fluid beta-amyloid(1-42) and tau levels (two established Alzheimer markers) and apolipoprotein E genotypes in patients with Alzheimer disease. Cerebrospinal fluid measures of beta-amyloid(1-42 )and tau differ between patients with Alzheimer disease and elderly normal controls. The effect of time and apolipoprotein E genotype on these biomarkers continues to be elucidated. We assessed cerebrospinal beta-amyloid(1-42) and tau in 20 mild-to-moderate Alzheimer disease patients, 11 apolipoprotein E4-positive and 9 apolipoprotein E4-negative, over a mean time of 3.8 years (range 1-11.1 years). Over the period measured, cerebrospinal fluid beta-amyloid(1-42) levels were lower in apolipoprotein E4-positive compared to apolipoprotein E4-negative patients, and the levels decreased over time. Tau levels were stable over time and did not show an effect of apolipoprotein E allele. Although these results were based on a limited clinical sample, the further decrease in cerebrospinal fluid beta-amyloid(1-42) (i.e., more abnormal) combined with the cerebrospinal fluid tau stability over a mean period of almost four years suggests that beta-amyloid(1-42) and tau maintain their potential usefulness as diagnostic biomarkers over time. These findings should be taken into account if cerebrospinal fluid beta-amyloid(1-42) and tau are used as measures of treatment response.